On Friday, January 15, 2021 4:09:05 PM CET Pierre-Louis Bossart wrote:
>
> >>> In addition to what Hans asked for, can you please build the kernel with the
> >>> debug patch below applied (instead of the Hans' debug patch), try to boot
> >>> the affected machine with it and see what is missing with respect to booting
> >>> the kernel with the two problematic commits reverted?
> >>
> >> Sorry, not following. Are you asking me to apply the patch below as well
> >> as revert the two problematic commits? Or just the patch below?
> >
> > Just the patch below.
> >
> >> the boot process is stuck without the reverts and I don't have a serial link to
> >> see what happens (closed form-factor).
> >
> > The point is that the patch below may unstuck it, in which case it
> > should be possible to find out what is missing with respect to the
> > full successful boot.
>
> No luck. I tried twice with the patch below only, and the device is
> still stuck after the 'Loading initial ramdisk ...'.
Thanks!
This means that skipping the enumeration of a certain device alone is
problematic which is a surprise of sorts.
Let's see what device that may be.
Because the machine booted with the debug patch from Hans, it should also boot
with the one below, so please do that and send the output of
$ dmesg | grep Dependencies
---
drivers/acpi/scan.c | 9 +++------
1 file changed, 3 insertions(+), 6 deletions(-)
Index: linux-pm/drivers/acpi/scan.c
===================================================================
--- linux-pm.orig/drivers/acpi/scan.c
+++ linux-pm/drivers/acpi/scan.c
@@ -1950,10 +1950,8 @@ static acpi_status acpi_bus_check_add(ac
if (type == ACPI_BUS_TYPE_DEVICE && check_dep) {
u32 count = acpi_scan_check_dep(handle);
/* Bail out if the number of recorded dependencies is not 0. */
- if (count > 0) {
- acpi_bus_scan_second_pass = true;
- return AE_CTRL_DEPTH;
- }
+ if (count > 0)
+ acpi_handle_info(handle, "Dependencies found\n");
}
acpi_add_single_object(&device, handle, type, sta);
@@ -1961,8 +1959,7 @@ static acpi_status acpi_bus_check_add(ac
return AE_CTRL_DEPTH;
acpi_scan_init_hotplug(device);
- if (!check_dep)
- acpi_scan_dep_init(device);
+ acpi_scan_dep_init(device);
out:
if (!*adev_p)
